It is estimated that 4.2% of the U.S. population is dependent on marijuana. Among those seeking treatment for their marijuana use, only a small percentage are able to achieve abstinence. During our previous funding period, we developed a laboratory model that identified medications that attenuate symptoms of marijuana withdrawal. The objective of this competing continuation is to refine our model to predict medications that decrease relapse to marijuana use, defined as the resumption of marijuana smoking after a period of abstinence. The underlying assumption of this proposal is that 1) dependence and withdrawal play a role in maintaining frequent marijuana use, and 2) attenuating either marijuana withdrawal or marijuana's positive reinforcing effects will decrease relapse. Aim #1. Develop a laboratory model of marijuana relapse. We hypothesize that abstinence will: (1) increase symptoms of withdrawal, (2) decrease the latency to self-administer marijuana, (3) increase the quantity of marijuana self-administered. Aim #2. Evaluate the ability of pharmacological interventions to decrease marijuana relapse by attenuating symptoms of marijuana withdrawal. Using this model, we will evaluate oral THC and nefazodone, which we have shown decrease symptoms of marijuana withdrawal. We hypothesize that both oral THC and nefazodone will: (1) dose-dependently increase the latency to self-administer marijuana, and (2) thus, dose-dependently decrease the quantity of marijuana self-administered following a period of marijuana abstinence. Aim #3. Evaluate the ability of pharmacological interventions to decrease marijuana use by attenuating marijuana's direct effects. We will evaluate the cannabinoid receptor antagonist, SR141716A. We hypothesize that SR will dose-dependently decrease the quantity of marijuana self-administered following a period of marijuana abstinence. Little is known about the factors contributing to the high relapse rates in marijuana treatment-seekers. The strength of this protocol lies in our utilization of a controlled laboratory setting to examine the interactive effects of medications on marijuana self-administration under conditions that model relapse. The data collected will provide information on the positive and negative reinforcing factors maintaining near-daily marijuana use, and will suggest more efficacious approaches to treating marijuana dependence.